Publish or Stay Behind and Perhaps Perish

Transcription

Publish or Stay Behind and Perhaps Perish
Soziale Systeme 11 (2005), Heft 1, S. 129-150
© Lucius & Lucius, Stuttgart
Wolff-Michael Roth
Publish or Stay Behind and Perhaps Perish:
Stability of Publication Practices in (Some) Social Sciences
Zusammenfassung: Obwohl neue technische Entwicklungen das schnelle und hinsichtlich der Länge problemlose Veröffentlichen ermöglichen, werden elektronische Medien
in manchen Wissenschaften nur langsam – wenn überhaupt – akzeptiert und benutzt.
Auf der Grundlage eines kulturhistorischen Ansatzes der dritten Generation argumentiere ich, dass sich die Stabilität von Veröffentlichungspraktiken (in Nordamerika) aus
der Rolle der Publikationen in der akademischen Laufbahn ergibt. Entscheidungen in
so unterschiedlichen Zusammenhängen wie Dauereinstellung, Beförderung, Gehalt,
Gehaltserhöhung, und Drittmittelerwerb hängen von der Veröffentlichungsliste ab, die
als eine Form der Objektivierung der Leistung eines Individuums verstanden wird. Die
Stabilität der Veröffentlichungspraktiken kann man daher als das Produkt von der hoch
vernetzten Natur akademischer Praktiken und Tätigkeitssystemen und der dialektischen Natur der Wissenschaftsgemeinden (communities of practice) verstehen, die sich
sowohl identisch reproduzieren (Stasis), als auch in neuen Formen produzieren. Dieses
Phänomen kann man zum Teil verstehen als das Bedürfnis eines Akademikers (einer
Akademikerin), zur Erhaltung der Wissensgemeinde durch Dienste beizutragen, die
den Entscheidungen über Dauereinstellung, Beförderung, und Gehalt Rechnung tragen.
I.
Introduction
A colleague working as science educator recently asked me about my tenure
as one of the co-editors of FQS: Forum Qualitative Sozialforschung/Forum Qualitative Social Research, a refereed online journal for qualitative research methods. He asked me in particular about the number of hits per month we
received to which I responded that in March 2004, for example, we have had
2,889,834 hits, 47,476 pdf file downloads, and 178,481 html file accesses and
that our newsletter goes out to over 4,000 subscribers (Mey/Mruck 2004). My
colleague then asked me what we had done to get these tremendously high
numbers. In our subsequent conversation, he articulated the concern that his
own discipline, science education, has seen only two electronic journals since
the Internet has made new forms of publishing possible, and both journals
(Electronic Journal of Science Education [http://unr.edu/homepage/jcannon/
ejse/], Electronic Journal of Literacy Through Science [http://sweeneyhall.sjsu.edu/
ejlts/]) are struggling, with respect to readership, contributions, and recogni-
130
Wolff-Michael Roth
tion.1 That is, although other journals in the field such as Science Education or
the Journal of Research in Science Teaching continue to attract authors, despite,
for example, their rejection rates of about 75 percent, other journals in the
field, and in particular the electronic journals are not doing well. My colleague
wanted to know whether I knew the reasons or had a hunch for this phenomenon, to which I responded with a hypothesis: The stability of publication
practices can be understood to arise from the highly interconnected nature of
practices (and activity systems) in the academy, and the dialectical nature of
any communities of practice, which reproduces itself in (nearly) identical ways
(stasis) as much as it produces itself in new forms. In this contribution, I articulate a theoretical framework based on networks of activity systems to discuss
concrete circumstances in which the stability of publication practices is continuously produced and reproduced. That is, I do not take stability for granted but
consider it as an achievement of networks of interacting activity systems. Taking
such a stance forces the analyst to articulate the processes that continuously
operate to stabilize the publication practices of a field, here, those that lead to
the virtual rejection of new forms of publication opportunities in fields such as
science education.
II.
Agency|Structure Dialectic
Many if not most social and psychological theories are deterministic and
therefore stand little chance to explain the variability of human (social)
actions. For example, socioeconomic factors are often cited as a cause of family
violence (e.g., Martin/Tsui/Maitra/Marinshaw 1999) or as substantial and
independent determinants of low school achievement (e.g., Payne/Biddle
1999). Such theories, however valuable to policy makers they may be, fail to
account for all those concrete cases in which people living in poverty do not
enact violence toward spouses, and where children living in poverty do in fact
show high achievement. Similarly, school achievement is frequently explained
in terms of causation: more intelligent individuals do better than less intelligent ones. However, I am not aware of nor could I find studies that correlate
measures of intelligence and other variables with scholarly productivity. Even
if such studies existed, they would not be able to explain every case of more or
less productive university faculty. In my research, generative theories based on
an agency|structure dialectic, and one of its particular realizations, third-generation activity theory have allowed me to understand and explain a wide variety of social and psychological phenomena without putting individual actors
1 As of April 30, 2004, the latter journal is listed in the Directory of Open Access Journals
(http://www.doaj.org/), but not the former.
Publish or Stay Behind and Perhaps Perish
131
into deterministic straightjackets of social or psychological factors (e.g., Roth
2003). In the following, I sketch these two aspects of my approach drawing on
examples pertinent to this special issue, that is, patterns of publication in the
social sciences.
Generative theories in cultural sociology and social psychology are grounded
in an agency|structure dialectic (e.g., Sewell 1992). Structure has two aspects
that bear on social action. On the one hand, there are the objectively experienced sociomaterial2 structures that are resources on which human beings
draw in their concrete actions – authors use »the literature« or a »word processor« as material resources for composing a research article and may interact
with journal editors in particular ways, »because« of the latters’positions in the
scientific community. On the other hand, schemas are embodied (»mental«)
structures that allow social actors to perceive and act toward material structures – authors at different points in their careers and with different levels of
experience view the same literature, materially embodied in journals and
books, in different ways. In a generative model of human social action, the two
forms of structure (resources, schemas) stand in a dialectic relation rather than
constituting a perfect homology, for the later leads to a deterministic model in
which no change and (individual and cultural) development is possible.3 From
both cultural-historical and ontogenetic perspectives, sociomaterial resources
(in the way they are perceived) and schemas (structures that generate perception and outward action) emerge together as human beings act(ed) in the
world (Roth 2003). Thus new aspects of sociomaterial practices, resources, and
schemas are always produced at the same time that other aspects are reproduced – even in the most mundane job at the assembly line, patterned actions
(i.e., social practices) change both at the individual and collective levels in the
course of praxis.
To articulate and theorize the structural aspects of human activities, some theorists (e.g., Engeström 1999) use a heuristic (Figure 1) that highlights the
2 I use the notion of »sociomaterial« structure, because, with Durkheim (1895, 45), I hold that
the concreteness of social facts – always encountered in their concrete material detail of
human interaction—is sociology’s fundamental phenomenon; at the same time, the way in
which material structures are apparent to the individual social agent is the result of social
mediation (e.g., Mikhailov 1980). The social and material mutually presuppose one another,
captured in the binding of the two adjectives into one.
3 Bourdieu frequently has been critiqued for the duality of his habitus-field pair of structures,
which leads to the exact reproduction of existing structures, so that a system and its social
actors is captured in stasis. Such criticisms cite passages such as, »As an acquired system of
generative schemes objectively adjusted to the particular conditions in which it is constituted,
the habitus engenders all the thoughts, all the perceptions, and all the actions consistent with
those conditions and no others« (Bourdieu 1977, 95). Bourdieu, however, repeatedly rejected
deterministic interpretations of his model and emphasized instead the generative nature of
his model: »The notion of habitus accounts for the fact that social agents are neither particles
of matter determined by external causes, nor little monads guided solely by internal reasons,
executing a sort of perfectly rational internal program of action« (Bourdieu / Wacquant 1992,
136).
132
Wolff-Michael Roth
mediational role of production means, rules, community, and division of labor
that operate within activity, the irreducible analytic unit of analysis. This
heuristic was developed presupposing an agency|structure dialectic though
articulated in terms of a subject|object dialectic. I explain the difference drawing on an example. Researching is a pervasive and archetypical activity in
academe: Its results, published papers, are so crucial to academic careers that
all new scholars are told by their elder peers, »publish or perish.« After identifying the nature of the activity to be researched, in my case researching and
publishing, social scientists proceed to identify possible structures – the most
obvious ones being the human subject and its object of activity, two mutually
presupposing (dialectic) entities. What the object of activity is can be established only with respect to the subject of activity, and the nature of the subject
can only be established with respect to the object of activity. It is not surprising, therefore, that the object has been conceived of as appearing twice, as
both material entity and (the subject’s) vision (Leont’ev 1978), a formulation
that reproduces the dialectic nature of structure. For the same reason, though
hardly ever articulated, the subject appears twice, as material body (»flesh«)
and person (Roth in press).
Figure 1. Heuristic for finding structure in human activities – treated as irreducible units of analysis – using researching in social science as an example.
The community defines the nature of the object (sometimes also motive«) in
activity and is itself defined by the object: object and community arise together
in the course of history and as a consequence of a division of labor. Thus, theoretician and researcher communities in education (Roth 2002a) or architecture
(Turnbull 1993) emerged only after the relevant professional (practitioner)
Publish or Stay Behind and Perhaps Perish
133
communities of practice had established and as the result of a split from these
communities of practice. In the same way that the tools available to researchers
mediate the shape observed activities take (a researcher who knows only
ANOVA will unlikely conduct an ethnographic study), rules mediate the
processes and products of research and writing (research methodology
describes, for example, the level of involvement between researcher and
researched).4 Most important to a dialectical activity theory is its cultural historical dimension, which is not captured in the static heuristic of Figure 1.
Activities continuously unfold, entraining changes in their constitutive elements, so that neither activity systems as wholes nor their constitutive aspects
can be understood outside the cultural-historical context in which they take
place and which they shape in return. Recent articulations of the theory therefore refer to this approach as cultural-historical activity theory (CHAT). In
doing research in the ways that other members recognize and accept it, the
individual subject not only reproduces practices and the community they constitute, but also produces and reproduces itself as an, in its concrete practices
recognizable member of this community. The individual subject and its collective culture, expressed in but not entirely represented by the community, stand
in a dialectical relationship. The individual always concretely realizes possible
actions that exist at the collective level, and in concretely realizing culture both
reproduce it and produce new variants of it, which leads to the development of
cultural possibilities.
Although activities constitute the unit of analysis, there are two further levels
of events that need to be distinguished, or, in other words, lead to structure:
actions, which are directed toward goals formulated by individual subjects and
operations, which are unconscious and occur in response to current conditions. Thus, whereas researching has a collective motive, writing an introductory
paragraph for an article about the stability of publication practices were clearly
enacted and pursued goals articulated by this author. Activities and the concrete actions that constitute them presuppose one another: activities are concretely realized by embodied actions in the pursuit of specific goals, but these
goals and actions are brought forth only for realizing a specific activity. The
sense of individual actions arises from their relations to specific activities: asking a question has a different sense when it is done to realize a research project
then when it realizes a contribution to the nightly news. Actions themselves
are constituted by unconscious operations: in the process of writing an introductory paragraph for this research article, specific words arise in response to
the current state of a sentence but I do not control the emergence of the words
4 I do not conceive of rules as determinants of action, but as resources to situated action. What
the relationship of established rules and situated actions is an empirical matter that can be
established only a posteriori (Suchman 1987).
134
Wolff-Michael Roth
themselves. My fingers on the keyboard produce these sentences, but I do not
consciously control their movement – apparent in my consciousness are shape
and content of the sentences that appear on the screen. Actions and the operations that constitute them also presuppose one another: properly sequenced
operations concretely realize an action, but the current state of the action conditions their emergence – I write sentences before having composed them in
my head, but the words that appear on the screen are not arbitrary and stand
in more or less ordered relationships to what is already there. Because actions
draw on embodied operations, which take considerable time periods to
become established, (cognitive) practices are durable simply because of the
investments involved in changing them.
III. Third-Generation Cultural Historical Activity Theory
Generative agency|structure dialectic and cultural historical activity theory are
not yet sufficient to explain phenomena at the collective (societal) level: activity systems do not exist in a vacuum, as they were treated in second-generation activity theory.5 They interact with, and are constitutive elements of, networks of activity systems. For example, the activity system as concretely
realized by the professor in Figure 1 receives rules and instruments from other
activity systems (e.g., American Psychological Association, tests, computers,
software), and produces outcomes for certain other activity systems (e.g.,
research participants, such as school districts) (see Figure 2). Thus, people and
artifacts (instruments, tools, reports) move from activity systems into other
activity system. This flow constitutes a heterogeneous network of otherwise
often independently functioning activity systems – professors doing their
research and publish are seldom conscious of the activities in the offices of the
American Psychological Association. The stability of networks of activity
results from this flow as well as from the durability of concrete people and
material things (Latour 1987), because they are now constitutive elements not
merely of one but in fact of multiple activity systems.
The movements of people and things from one activity system into another
are not a sufficient explanation for surprising events and changes in the activity of interest. The foreign entities are first appropriated by the activity system,
are modified, and made integral and in-dissociable part of it. At the same time,
this introduction can give rise to tensions, antinomies, contradictions, and
5 First-generation activity theory (e.g., Vygotsky 1978) focused almost exclusively on the toolmediated nature of activity, but did not theorize the dialectical relation that co-constitutes
individual and collective; second-generation activity theory focused on how a collectivity
mediates the object of activity, but did not concern itself with the interrelation of multiple
activity systems.
Publish or Stay Behind and Perhaps Perish
135
breakdowns: identifying, dealing with, and working through these, the activity
inhibiting situations leads to change. If we were interested in changing an
activity system, we would attempt to locate the contradictions, antinomies,
breakdowns, or tensions and, in a collective fashion involving other members
of the system, remove the problematic aspects. Because human subjects are
constitutive part of their activities, they have the power to produce change all
the while reproducing other aspects of the system.
Figure 2. Social phenomena are best explained in terms of the interaction of multiple activity systems, which occurs by means of a flow of people and things. The stability of the resulting networks
is proportional to the investments and consequences to the entire network when the flow is stopped.
Using cultural-historical activity theory in its third-generation form allows us
to demystify processes hidden from many researchers, and articulate the concrete actions that bring about the outcomes that the researchers actually come
to see, for instance, the decision letter from the editor to the author. In my
experience, many young scholars who receive rejection letters see these as
objective reflections of their work rather than as the results of processes in
which real people are involved, those with whom they interact, for example, at
conferences, and whose work they judge as reviewers. The third-generation
CHAT approach no longer articulates activities one at a time, but as simultaneous and interdependent activities that form a network. Second, third-generation CHAT encourages us to study the processes (always accomplished in concrete human actions) that bring about, for example, the entire review of a
submitted manuscript leading the production of a rejection or acceptance letter. In the following, I use third-generation cultural-historical activity theory to
articulate some aspects of scholarly activity that contributes to the stability of
publication practices in the social sciences, supported with evidence from the
discipline of science education.
136
Wolff-Michael Roth
IV. Publish or Perish
»Why,« to return to my colleague’s question, »do new (electronic) journals have
such a hard time to establish themselves [in some disciplines]?« The diction
»publish or perish« encourages us to study two potential sites that contribute to
the stability of publication practices: (a) the research-writing-submission-publication axis of a faculty member’s work, the results of which are inscribed in
and represented by the cumulative curriculum vitae (CVs); and (b) the – in
North American universities pervasive – evaluation of the faculty member by
salary, tenure and promotion committees or deans of faculty based on submitted CVs (Figure 3). At least in the Canadian situation, the research record as
inscribed in a CV is also an important element in the evaluation of research
grant applications to the major research funding source, the Social Sciences and
Humanities Research Council of Canada, where, for a regular scholar, 60 percent
of a proposal evaluation pertains to the CV and 40 percent to the project itself.
The publication record as embodied in a CV therefore constitutes an important
resource in garnering funding for future research. Investments in the resulting
history and network of activities – existing in the form of people, materials, and
Figure 3. A temporal look at an untenured faculty member’s recurrent research activities, which,
ideally, result in relevant publications that are subsequently represented as entries in the
researcher’s curriculum vitae (CV). During tenure and promotion decisions, the CV stands in for
(represents) the researcher.
Publish or Stay Behind and Perhaps Perish
137
practices – contribute to stabilizing particular genres and journals and make it
difficult for change to occur. Because of the linkage between researching-publishing, on the one hand, and career progress (survive or perish), on the other,
the motive (object) for engaging in research may actually be the building of a
CV to jump the tenure and promotion hurdle rather than the generation of new
knowledge.
IV.1 On the Perils of Not-Publishing
As a general rule across North America and across disciplines, »one or two
articles published in highly prestigious journals might win the author tenure
or a healthy pay raise, more articles published in less prestigious journals
might be needed to secure the same rewards« (Korobkin 1999, 858). »Producing enough« to make the tenure hurdle is the principal motive for much
research in North American universities, and is associated with considerable
anxiety on the part of new scholars. That is, the evaluation of productivity is
not pro forma, a done deal, but a serious activity with consequences for the
individual involved – requests for tenure, promotion, salary increases, or
research funding are rejected rather than automatically granted. Over the past
dozen years, I frequently participated on committees that evaluated faculty
members using their curriculum vitae as the main source of evidence. These
committees make decisions about tenure, promotion, salary increases, or
awards. Each of these situations can be considered as a concrete realization of
a form of activity, with the committee as the (collective) subject, the particular
CVs as the object (motive), and the production of a (written) evaluation as the
outcome (Figures 1, 3). In all of these committees, the value of contributions a
researcher has made to the scholarly literature and community is, to a great
extent, established in terms of the number, length, and quality of articles published.6 Whereas the number and length of articles – very short ones, editorials, or comments and criticisms count less than regular »full-length« articles –
can easily be established by inspecting the CV, the quality of journals is less
readily established, the outcomes depending on the committee composition.
Here, committees draw on a variety of means to assist them, including journal
rankings, impact, usefulness, and qualitative »tiers,« in the production of an
evaluation and decision, which is formalized in a letter that goes back to the
applicant.7 Cultural-historical activity theory allows us to articulate the various
elements that enter the decision-making process, and therefore they are integral to the outcomes. The evaluation articulated in the letter to the candidate
6 Elsewhere I provided a detailed analysis and modeling of the inner workings of committees in
funding agencies, and the vagaries that these may lead to (Roth 2002c).
7 This letter is another material entity that constitutes or contributes to constituting the flowing
matter that makes and stabilizes networks of different activity systems (Roth 2002b).
138
Wolff-Michael Roth
and dean is as much a reflection of the committee process as it is a reflection of
the candidates and their CVs.
In all of the committee meetings I attended, publications that appear in »flagship journals« have been rated considerably higher and counted positively in
the evaluation, whereas »second [third] tier journals« not only may count less
but also, in some cases, may count against the person. Thus, to a North American researcher in science education the Journal of Research in Science Teaching
and Science Education constitute first tier journals, whereas the International
Journal of Science Education and Research in Science Education might constitute
tier two journals, and so forth. To establish the quality of the journals, that is, to
attribute them to »first tier,« »second tier,« or »third tier,« committee members
may draw on and articulate different forms of evidence. For example, many
members and external evaluators draw on their own sense and experience to
establish the impact a journal has on the field. Such evidence is very personal,
and some individuals, as I could ascertain on two committees in the months
prior to writing this article, may therefore value journals with a regional or
national orientation as highly as those journals – unknown to them – that
truly have an international orientation and readership. Other members, on the
other hand, may use »objective measures« such as rejection rates (obtained
from editors), journal impact (as provided by Journal Citation Reports®), journal
rankings (as a result of ranking using impact ratings), or relative journal rank
within discipline8 as indicators. Judgments based on influence, significance, or
importance of research publications could be used, but they require qualitative
analysis by experts in the field, a process often too time-consuming and
expense to be feasible (Garfield/Welljams-Doroof 1992). Tracking how often
papers are cited, citation databases constitute tools for indicating impact of
primary research papers and, after aggregation, can indicate the relative
impact of individuals and journals as well as larger units such as departments,
disciplines, fields, and the science as a whole.
Committees consider the extent to which the applicants contributed to the
community of which they are constitutive members. Service as reviewer,
member of editorial board, associate editor, or editor are considered to be
legitimate contributions to the field, increasing in value in the order listed
here. Researchers seeking tenure or new jobs therefore ought to participate in
the discipline by serving as a reviewer and, during later stages of their career,
as members of editorial boards, associate editors or editors. The value of the
8 Relative journal ranking in the discipline r can be calculated according to the equation r = 1 (n - 1) / N, where n is the descending rank number and N the total number of journals in the
discipline (Popescu / Ganciu / Penache / Penache 1998). That is, the Journal of Research in Science
Teaching (JRST) was ranked 8th of 102 journals, it would have a value of r = 1 - (8 - 1) / 102 =
0.93, which means that 93 percent of journals in the field have a rank and corresponding
impact factor lower than JRST.
Publish or Stay Behind and Perhaps Perish
139
service, as considered by the committee, again depends on the journal with
which the person associates itself, which can, in disciplines such as law, be
critical to obtaining coveted jobs in one of the major law faculties (Korobkin
1999). It may therefore come as little surprise that a substantial number of the
editorial board members of the Journal of Research in Science Teaching are
untenured assistant professors and recently appointed associate professors. To
review and even more so to be a member of an editorial board, one needs to
have published in the journal.
Funding records also contribute as means to committee engaged in the evaluation of scholars, and deans of faculty impress upon new members the importance of getting grants to their careers. However, getting grants heavily
depends, at least in Canada, on publication records. In the Social Sciences and
Humanities Research Council of Canada, the proposals of new scholars (first 5
years) are judged according to a 40:60 or 60:40 percent split for record and
quality of project (whichever gives a higher score), whereas regular scholars
are judged according to the mentioned 60:40 record-project weighting. Thus,
although it is desirable to be able to list research grants, getting them is itself a
function of prior research publications as represented in CVs. A frequent
advice to new scholars therefore consists in statements such as »Getting on
the funding train early or missing it for the entirety of a career.« Furthermore,
national impact is valued higher than a regional one, and international impact
higher than national, each distinction defining a ten-percent cluster. This, in
turn, favors publications in international and high-impact journals, which
receive higher evaluations than national or regional journals and higher evaluations than journals that are not ranked at all.
As a result of their activities related to researching, publishing, doing service to
journals, and getting grants, faculty members increase the numbers and
weightiness of lines in their CVs.9 The CVs not only are representations of
researchers’accomplishments, but also becomes their representatives (spokesperson) that are interrogated by and provide answers to the respective committees.10 In decisions that affect the careers of researchers, members of the
relevant committees draw on a variety of data, all of which are the outcomes of
activity systems directly or indirectly linked to doing research (Figure 3).
Researching contributes to publishing, which itself contributes to being cited,
9 An Australian colleague takes an entirely pragmatic perspective on research activity, counting
all returns in terms of their value to the CV. Thus, he greets every publication with the comment, »Another line on my CV.« Everything he does seems to be connected to adding lines to
his CV, and I only hear from him when another line is at stake. As the editor of a journal, he
even regards »weaker« articles as counting against not only the journal but against him and
his record.
10 This articulation of a representation as a spokesperson for the person it represents is consistent with an actor network approach, which treats human and non-human actors symmetrically (Roth / McGinn 1998).
140
Wolff-Michael Roth
serving in different functions in journal operations, and getting grants. That is,
publication records weigh heavily in direct and indirect ways on scholarly
careers. To make it through the various decision processes where they have to
submit their CVs to represent themselves, faculty members at all stages of
their career orient towards (are motivated by) the various elements that enter
the considerations. That is, the motives concretely realized in the researching
and publishing activities (Figures 1-3) are mediated by the means that enter
decision making in other activity systems, which, in turn, draw on objects,
people, outcomes of networks of activities that have formed around publishing.
In this mix of activities publishing and participating in maintaining new journals is therefore a risky business for new faculty members, who, in science
education for example, do much of the publishing, reviewing, and editorial
board service. Publishing in one of the new electronic journals or doing service
to maintain it alive may not only count little but even count against the faculty
member.
IV.2 On Citing and Being Cited
Authors and the journals in which they (aspire to) publish are part of selfreproducing and self-stabilizing networks of activity systems. Journals exist in
a network of activities that involve materials, people, and practices that contribute to stabilizing the networks (Figure 3): as soon as the relation between
one activity system and another is held for an amount of time, it generates
effects of stability within and across systems as well as the conditions of power
(Law 1991). The flow of objects between systems links different activities
(authoring, editing, reviewing, and publishing) and thereby contributes to the
performance of relationships, such as editor-author (acknowledgment of
receipt, decision letter, letter of acceptance), reviewer-author (review), editorreviewer (request to review, manuscript, review), editor-publisher (manuscripts), publisher-author (proofs), and so forth. People are involved in all of
these relations, not only those already listed, but also support personnel
within the different activity systems including secretaries (making copies), editorial assistants (managing manuscripts and databases), type-setters, managing editors, printers, and web administrators (for electronic publishing).
Peer review involves the creation and movement of many documents, only
some of which the authors get to see. In addition to the manuscript, there may
be reviewer rating forms, letters from the editor to the reviewers, reviews with
comments to the editor and comments to the authors, editorial decision letters, electronic or paper databases for tracking documents, statistical information about manuscript flow, cover letters specifying changes in revisions,
acceptance letters, marked up manuscripts, galley proofs, and reprints. All of
Publish or Stay Behind and Perhaps Perish
141
these documents constitute durable material objects; the flow of such durable
objects stabilizes networks of activity systems by reproducing and therefore
confirming and reifying existing relations (Roth 2002b).
This stability arises from the flow of these intermediaries, a current that continues even if some of the actors disappear from the network because they opt
out or are removed (e.g., when tenure is denied). Manuscripts, reviews, decision letters, and CVs are intermediaries that function like any other intermediaries (currencies) that become the lifeblood of a system, seemingly indispensable; this fluid constitutes the social topology of a discipline. This fluid
produces a level of stability that makes any change effort difficult. All of these
events and processes contribute to the stability of form and content of publications even in the apparently most solitary form of scholarly pursuits, doing
and writing philosophy (Cronin 2004).
Journals also face the potential of perishing; they have to contribute to staying
in the field, which they do by contributing to the stability of the networks in
which they are involved. Journals are not things that stand on their own but
have their own context, that is, they are judged in the context of other journals
in terms of comparison of readership, distribution, importance to the field,
quality of the articles published, nature of the articles published, and so forth.
There is a network within which journals operate and are caught up, one that
is established by the ranking procedures of such institutions as the Social Sciences Citation Index. A criterion often invoked for tenure and promotion purposes is the impact of the author on the field. One measure of impact is the
citation statistic, which to a large extend depends on the spread of the journal
rather than the quality of the article. An article published in the Educational
Researcher, which is automatically mailed to more than 10,000 members of the
American Educational Research Association, has a higher spread and therefore likelihood to be cited than an article by the same author published in the
Journal of Research in Science Teaching, also sent automatically to the roughly
1,500 members of the National Association for Research in Science Teaching
(despite its name, an international organization). An article in Review of Educational Research with a large subscription base has a higher potential to be cited
than if it was published in the Journal of Research in Science Teaching with a
smaller distribution. At the same time, journals with a small number of citable
articles and issues per year, such as Review of Educational Research (1998-2002,
Nav = 14) or Journal of the Learning Sciences (1998-2002, Nav = 11), tend to be
found among the highest ranked journals in the field. Given the way citation
rankings are calculated, one or two well-published authors can significantly
affect the citation index and therefore the rankings of these journals.
Citation studies are often undertaken to evaluate faculty members and institutions in addition to studying structure and development of a field or cited
motivation (Rousseau/Zuccala 2004). Such studies show how different social
142
Wolff-Michael Roth
actors operating in the same or different activity systems generate resources
that tend to produce and reproduce stability of publication practices. These
social actors – that is, constituents of the subject|object dialectic in various
activity systems – include authors, journals, institutes, countries, scientific
journals, and combinations thereof as factors.
Authors contribute to the stability of publication activities by producing and
reproducing citation networks and collective interests. To get their work
accepted in the »flagship« journals, authors have to show that their research
object is of collective interest – pursues a collective motive – that their research
methods conforms with standard methodologies (rules), draw on the wellknown papers and authors as resources (tools) to support their arguments,
and that they use standard formats (genres) as means to communicate their
ideas. They combine these with »original data,« to create research narratives
that are, in some ways, reflective of the setting in which the protagonists (and
authors) have lived. They draw on other actors such as previously published
reports, common knowledge, and established scientific processes and experimental procedures to construct the reasonableness of their research question
and experimental design. Common knowledge and widely accepted facts,
concepts, and theories are more powerful supporters in an author’s scheme
than other yet-unconfirmed research findings; articles by »authorities« are
more powerful allies than articles by largely unknown researchers. Leading
authorities and »more important« articles are those published in flagship journals. Thus, my experience serving as associate editor or editorial board member shows that citing an author in a second- or third-tier journal (e.g., School
Science and Mathematics) or an online publication (e.g., Electronic Journal of Science Education) is taken as weak support to the statement a science education
author makes than if the citation is to a leading author and journal. The
authors have to anticipate the possible ways in which editors and reviewers
might read their manuscript, and they use these anticipations in shaping their
manuscript. That is, a manuscript inherently becomes a cultural object not only
because the language used »exists asymmetrically, always for the other, from
the other, kept by the other. Coming from the other, remaining with the other,
and returning to the other« (Derrida 1998, 40) but also because genre, citations, referencing, and other aspects of the text as a whole can be recognized
by other members of the discipline as a concrete realization of discipline-specific cultural possibilities. In producing manuscripts that are recognized by
others in this way, authors mainly reproduce cultural possibilities and shy
away from producing new texts and possibilities.
Citing the work of others constructs links between authors and journals. Thus,
each time authors cite articles that have been published in a leading journal,
they also contribute to stabilizing the network of journals within which they
exists and to reaffirming and reifying the status of the journal and the power of
Publish or Stay Behind and Perhaps Perish
143
the editor. Each citation – which is a material sign that points to and has been
occasioned by an article, and therefore part of the material flow – contributes
to the maintenance of the existing network. Each citation even contributes to
the maintenance of journals to which authors do not submit their work (Roth
2002b). In the same way, each citation of an author’s work by other authors
contributes to the calculation of the citation index and impact rating. Again,
they thereby contribut to stabilizing the journal and its editor – as citing author
and as cited author.
Many authors – in part depending on the type of institution, college, teaching
university, research university (»Research I«) or comprehensive university –
attempt to publish in the »flagship journals« of their field, in fact, they have to
publish in these journals if they want to make the tenure barrier. However, one
might ask, »What constitutes a flagship journal, and why do others belong to
the ›second tier‹, ›third tier‹, and so on?«11
Journals that are already ranked high, on subjective-qualitative or quantitative
basis, receive more submissions and are more desirable outlets for scholarly
activity than others that are ranked lower. In science education, the two »leading« journals Journal of Research in Science Teaching (JRST) and Science Education
are also those that rank highest in the ISI citation index according to the journal impact factor – often making it among the top ten of the more than 100
education journals ranked.12 Both journals reject 75 percent or more of submitted articles (for JRST, between 220 and 250), and therefore tend to feature
articles of only a small fraction of authors in the worldwide community of science education. These rejected articles are frequently submitted with little
change to the next journal on the list of rankings.13 But now, there are many
print journals to which an author in science education may submit. It may
therefore come as little surprise that (a) the journals in the field publish the
same kind of articles and (b) the third journal in the rankings, the International
Journal of Science Education, has increased its numbers from six per year to 12,
11 In the field of law, for example, five tiers of authors and journals have been identified
(Korobkin 1999).
12 The journal impact factor is a measure of the average citation frequency for a specific citable
item (article, review, letter, discovery account, note, and abstract) in a specific journal during a
specific year or period (Garfield 1979). Thus, the impact factor for a specific journal in 2004 is
calculated (total citations in 2002 and 2003) / (total citable articles in 2002 and 2003). This
means that journals with a limited number of articles may easily become those with the highest impact factor. The »leading« journal in education in terms of cumulative rankings over 10year spans, Review of Educational Research, publishes less than 15 articles a year and has an
impact factor of about 3. That is, a small number of authors may be responsible for pushing
the journal to the top.
13 Having served on many journals concurrently, I have been asked repeatedly to review the
same manuscript for different journals. Old-timers in science education provide newcomers
with the advice of sending a manuscript first to the top journal, Journal of Research in Science
Teaching, and then, if it is rejected, to »go down the list,« where it is implied that the list is
rank-ordered from most-desirable (»best«) to least-desirable journal.
144
Wolff-Michael Roth
and most recently to 15. At the same time, the Electronic Journal of Science Education has had insufficient submissions and therefore not published a single
issue between December 2002 and the time of this writing (its home page provides as reason, »EJSE is experiencing a major backlog in its publication
schedule. Efforts are underway to bring the journal up to date.«); the Electronic
Journal of Literacy through Science has not received a substantial number of
manuscript and published only 8 articles over its two years of existence.
Citations are an important means of producing and reproducing scholarly
communities and sub-communities networks, and they can be used to map
the intellectual content of field and demarcate their (porous) boundaries
(Hummon«Doreian 1989). But there is more to citations. Authors want to
publish in flagship journals, but publishers want to maintain their advantage
(sales, readership): they have to maintain their staff to be able to publish their
palette of journals. Both are interested in keeping the journal rankings and
perceived importance high. Publishers actively contribute to the role of citation indices and journal rankings by using them in their advertisements and in
the construction of team spirit in editorial board meetings. Thus SAGE Publications feature the rankings of several journal on top of the journal home page;
the Wiley Publishers representative begins the editorial board meeting of the
Journal of Research in Science Teaching with the citation information from the
past several years; and an editor of Science Education pointed out in an editorial
board meeting out how far the journal has climbed in the ranking, a fact that
he attributes to the editorial board and his own editorship (»I wanted to let
you know that most recent Journal Citation Report showed that out of 101
journals published in educational research, Science Education ranks #10. Everyone affiliated with the journal should be proud of this accomplishment« [email
July 3, 2001]). It is therefore of interest to companies to provide editors with
support (e.g., paying for a [graduate] editorial assistant); and editors want to
maintain the edge of their journal to warrant and perhaps increase the support
they receive, which allows them to increase the number of issues that they
publish per year. Thus, over the past 15 years, the three major journals in the
science education discipline have increased the number of issues or pages
published (Table 1). This support, »hard currency,« constitutes another material object that stabilizes the relationship between editors and publisher. By
increasing space, science education journals have attracted authors and readers, but the manuscript rejection rates have stayed constant.
Journal rankings may mediate the nature of scholarship published, the stabilization of journal rankings. Thus, rankings can create incentives for journal
editors to select certain types of manuscripts and the selection of certain types
of manuscripts can create incentives for authors seeking publication in those
journals to produce those types of manuscripts (Korobkin 1999). The practice
of ranking journals constitutes a form of feedback by increasing perceived
Publish or Stay Behind and Perhaps Perish
145
importance. Thus, four of the five leading journals in clinical medicine
increased their relative citation impact on the field (Garfield/Welljams-Doroof
1992), stabilizing these journals against the potential impact of other publication outlets.
Table 1. Number of issues and pages published in three science education journals in 1989
and 2003
Year
Journal
Journal of Research in
Science Teaching
Science Education1
International Journal of
Science Education2
1989
Number of Number of
Issues
Pages
2003
Number of Number of
Issues
Pages
9
5
784
593
10
6
1114
918
3
362
12
1544
Note 1: There were 6 issues, but at the time one issue was entirely devoted to a summary of
the literature. In 1989, there were 141 pages in the summary issue.
Note 2: Starting 2004, this journal publishes 15 issues/year
Journals tend to support particular formats, genres, and research methods.
Perceived tendencies for formats, genres, and research methods become
resources for authors to shape their own articles. If journal editors adjust their
article selection criteria in an attempt to increase their relative prestige,
authors will adjust the content, format, and genre of their articles because they
wish to place their articles in the highly ranked journals. As the leading
authors, editors, and editorial board members all have emerged from the same
community of practice, the currently leading journals in science education differ very little from one another. This point is part of the argument for a new
journal made in a recent proposal for a new and different journal in the field:
Of concern to the editors of the proposed journal is that those
reviewers and journal editors who tend to reify methodologies
grounded in positivism and conceptual change theory harshly treat
articles seeking to break the mold, such as the ones we would be
soliciting. Articles that do not conform to these methodological and
epistemological underpinnings are frequently rejected in the other
journals (Submission to Kluwer Academic Publishers).
This certainly contributes to the fact that most journal articles in the field are
not only highly uniform but also that the journals differ very little in the content and format of the articles they publish, though they differ in terms of the
quality standards that the field sets. Thus, the proposal suggested:
146
Wolff-Michael Roth
Although there are numerous journals in science education at the
moment, most are almost identical in what they publish and
encourage for submission. Leading journals such as Journal of Research
in Science Teaching, Science Education, International Journal of Science
Education, Research in Science Education, and Journal of Science and
Technological Education are virtually indistinguishable. Editors, editorial
boards and reviewers are selected to produce forms of scholarly
activity that are primarily grounded in traditional psychology and
inattentive to theories from cultural studies. There is a focus on
production with each journal seeking more pages per year and a press
on authors to write shorter articles (Submission to Kluwer Academic
Publishers).
These common formats are resources of limited range to new authors, who
come to produce manuscripts of a similar type and thereby reproduce the field
as it is. An author can begin submitting a manuscript to the most highly
ranked journal, and then, if necessary, move down the list without making any
changes in the text, until a journal actually accepts it. In summary, then, the
developments over the past two decades have seen an expansion of the traditionally dominant journals in science education and the few additions that use
the traditional print format are reproducing the practices of the existing ones.
On the other hand, despite the promises of open access and potential for wide
distribution made possible by online journals (e.g., Mey/Mruck 2004), this
new form of publishing has made extremely marginal inroads in science education.
IV.3 Where are the Others?
Against the domination of the three major journals in the discipline, other
print journals in science education find it difficult to exist, and the electronic
journals tend to have very little appeal. Among the journals, only Research in
Science Education has recently been entered in the ISI listing, after the Australasian Science Education Research Association sold it to Kluwer Academic
Publishers. Other journals have a more limited appeal, are listed in fewer
abstracting systems, and fail to attract manuscripts by the leading scholars in
the field. Among these journals feature the Journal of Science Teacher Education,
International Journal of Science and Mathematics Education, Science & Education,
and Journal of Science Education and Technology, Research in Science & Technological Education, School Science Review, or Studies in Science Education. Though
often aspiring and publishing international authors, other science education
journals have even more marginal existence, including the Journal of the Korean
Science Education, and journals published in a language other than English are
virtually not cited at all, including Didaskalia (France), Zeitschrift für Didaktik
Publish or Stay Behind and Perhaps Perish
147
der Naturwissenschaften (Germany), or Enseñanza de las Ciencias (Spain). All of
these journals constitute outlets for scholarly productivity, but, because they
do not contribute to the citation network, do not significantly enhance (North
American) authors’ credibility within the field nor do citations in these journals add to journals and authors in the »flagship journals.« Currently, electronic journals are considered to be »even lower on the totem pole« (Larry
Yore, personal communication, January 2004) in part because of perceptions
that it is easier to get published, that the review processes are not as rigorous
as in the leading journals, that it is not desirable to publish in the journals
given that the leading scholars do not publish there and likely do not read
these journals, and so forth. A number of leading scholars also suggest that
there are »too many journals in the field.« That is, the »flagship journals« are
part of a self-stabilizing network, which excludes others, an exclusion that is
co-produced by the failure to contribute to the stabilization of citation networks – a reverse statement of the function cited-citing pairs in constituting
and stabilizing scientific communities (Leydesdorff 1998). To make inroads,
electronic publications need to link with established networks, such as being
part of catalogues of abstracts and secondary distribution means such as
EBSCO Publishing (Mey/Mruck 2004). Established scholars do not or little
contribute to lower ranked journals; they also sit on tenure, promotion, salary,
and funding committees using journal rankings and qualitative indicators as
measures of productivity for their more junior colleagues. Senior scholars hold
editor and associate editor positions, making decisions about what kind of
articles to include or exclude – directly, by not entering an article in the review
process, or indirectly, by sending it to reviewers that will certainly reject the
piece.14 They contribute to producing the stability of the network by reproducing the existing status quo, making it difficult to impossible for new formats
(e-journals) and genres to establish themselves.
14 An example excerpt of such a letter states: »I have elected not to place this manuscript into
the review process because after a careful reading I find that 1) it does not explicitly deal with
issues of science education; 2) it does not extend our knowledge about teaching practices e.g., your theory of practice does not report new and novel perspectives that inform preservice
or inservice teacher education; and 3) it employs a format for the reporting of ethnographic
research which does not make obvious to the reader how you move from the evidence to the
results and conclusions reported« (Letter from the editor, MAR 01, 1996). As an example of
the second case, the then editor of Science Education sent an article entitled »Towards an
Anthropology of Graphing« to a cognitive psychologist, who stated that he did not understand anthropology, that there was too much anthropology in the article, and that it should be
rejected.
148
V.
Wolff-Michael Roth
Conclusion
In this article, I articulated a framework for analyzing the apparent stability of
publication practices, drawing in particular on example from one discipline,
science education. The dialectical agency|structure framework allows us to
understand and explain, why and how each author, even those who try but do
not get published in flagship journals, contribute resources to the networks
that tend to stabilize them. That is, with each failure to get a research manuscript through the review process, an author actively contributes to producing
the collective interest in writing for these journals and to the rejection statistics, both of which tend to stabilize these journals. The third-generation activity theory contributes to understanding how the movement of material objects
contributes to the production and reproduction of networks of activity systems, each of which requires exchanges with other systems to stabilize simultaneously the network and itself. Although the tenure, promotion, and salary
decisions, which contribute to the stability of the flagship journals, do not exist
in other countries, funding decisions are made on similar grounds, leading to
the stabilization of research and publication practices in these countries as
well. Although restricted in their influence on the literature in the lingua
franca language English, publications and grant proposals tend to lead to
characteristic stabilities by drawing on particular citation practices that produce and reproduce the status of journals, editors, and scholars.
Because of the interconnected nature of the activities and activity systems relative to publishing, service (journals), grantsmanship, and being successful in
getting research published reaps additional rewards, which accumulate to
make curriculum vitae that will stand scrutiny during subsequent evaluation
processes. That is, success breeds success not in a linear but in an exponential
fashion. Success, however, crucially depends on the journals in which a
scholar’s work gets published, which tends to stabilize existing rank orders,
making it difficult for new journals to establish themselves. This stabilizes
existing journals and journal hierarchies at the expense of new journals, even
those that provide opportunities for changing the field. In the field of science
education, however, there is an interesting development that future analyses
might elucidate. Despite their international appeal, the Journal of Research in
Science Teaching and Science Education tended to be more nationalist, because
review processes disfavor non-English speaking authors, both because of linguistic problems with the pieces and with cultural differences in format and
genre.15 The International Journal of Science Education, on the other hand,
seemed to have made a particular effort in attracting scholars from around the
15 Reviewing more than 50 manuscripts per year, for 30 or more journals in several discipline, I
have learned to identify the country or continent of origin of a manuscript based on type of
research, genre of writing, format of the manuscript, and language.
Publish or Stay Behind and Perhaps Perish
149
world. The International Journal of Science and Mathematics Education has been
provided with funds to support non-English speaking authors in the editing
stages of their manuscripts. Whether publication practices in the discipline will
change as a result, and whether there are similar or different trends for the two
journals (the former listed in ISI, the latter not listed) would allow for additional refinements in models concerning the stability (or instability) of publication practices, will be an empirical matter for future analyses.
Citation is an emerging (reflexive) practice, in that citations stabilize journals
and authors and keep others out of existing citation networks. As decisions
about authors are made with respect to their impact on the field, reflexively
determined by their ability to publish in the »flagship journals,« there is little
motivation to bring about and support forms of publication (i.e., e-journals)
that for a variety of reasons discussed here fail to contribute to citation networks. Citation analysis implies a reflexive theory. While such a theory has not
(yet) been articulated, the present article provides a glimpse at some of the
processes that stabilize journals, citations, and authors, in part because of the
role they play with respect to tenure, promotion, salary, and funding decisions.
References
Bourdieu, Pierre (1977): Outline of a Theory of Structure. Cambridge: Cambridge University
Press.
Bourdieu, Pierre/Wacquant, Loïc J.D. (1992): The Purpose of Reflexive Sociology (The
Chicago Workshop). Pp. 60-215 in: Pierre Bourdieu/Loïc J.D. Wacquant, An Invitation to
Reflexive Sociology. Chicago: University of Chicago Press.
Cronin, Blaise (2004): Bowling Alone Together: Academic Writing as Distributed Cognition.
Journal of the American Society for Information Science and Technology, 55, 557-560.
Derrida, Jacques (1998): Monolingualism of the Other; or, The Prosthesis of Origin. Stanford, CA: Stanford University Press.
Durkheim, Emile (1895): The Rules of the Sociological Method. New York: Free Press.
Engeström,Yrjö (1999): Activity Theory and Individual and Social Transformation. Pp. 19-38
in: Yrjö Engeström/Reijo Miettinen/Raija-Leena Punamäki (Eds.), Perspectives on Activity Theory. Cambridge: Cambridge University Press.
Garfield, Eugene (1979): Citation Indexing – Its Theory and Application in Science, Technology, and Humanities. Philadelphia: Institute for Scientific Information Press.
Garfield, Eugene/Welljams-Dorof, Alfred (1979): Citation Data: Their Use as Quantitative
Indicators for Science and Technology Evaluation and Policy-Making. Science and Public
Policy, 19, 321-327.
Hummon, Norman P./Doreian, Patrick (1989): Connectivity in a Citation Network: The
Development of DNA Theory. Social Networks, 11, 39-63.
Korobkin, Russell (1999): Ranking Journals: Some Thoughts on Theory and Methodology.
Florida State University Law Review, 26, 851-876.
Latour, Bruno (1987): Science in Action: How to Follow Scientists and Engineers through
Society. Milton Keynes: Open University Press.
Law, John (1991): Power, Discretion and Strategy. Pp. 165-191 in: John Law (Ed.), A Sociology of Monsters: Essays on Power, Technology and Domination. London: Routledge.
150
Wolff-Michael Roth
Leont’ev, Alexei N. (1978): Activity, Consciousness and Personality. Englewood Cliffs, CA:
Prentice Hall.
Leydesdorff, Loet (1998): Theories of Citation? Scientometrics 43, 5-25.
Martin, Sandra/Tsui, Amy/Maitra, Kuhu/Marinshaw, Ruth (1999): Domestic Violence in
Northern India. American Journal of Epidemiology 150, 416-427.
Mikhailov, Felix (1980): The Riddle of Self. Moscow: Progress.
Mey, Guenter/Mruck, Katja (2004). Open Access und elektronisches Publizieren: Das
Beispiel der Online-Zeitschrift FQS. In (Ed.), dvs. Hamburg: Edition Czwalina.
Payne, Kevin. J./Biddle, Bruce J. (1999): Poor School Funding, Child Poverty, and Mathematics Achievement. Educational Researcher 28 (6), 4-13.
Popescu, Ioan-Iovitz/Ganciu, Mihai/Penache, M. Cristina/Penache, Dan (1997): On the
Lavalette Ranking Law. Romanian Reports in Physics, 49, 3-27.
Roth, Wolff-Michael (2002a): Being and Becoming in the Classroom. Westport, CT: Ablex.
Roth, Wolff-Michael (2002b): Editorial Power/Authorial Suffering. Research in Science Education, 32, 215-240.
Roth, Wolff-Michael (2002c). Evaluation and Adjudication of Research Proposals: Vagaries
and Politics of Funding (92 paragraphs). Forum Qualitative Sozialforschung/Forum:
Qualitative Social Research, 3(4). http://www.qualitative-research.net/fqs-texte/3-02/302roth-e.htm
Roth, Wolff-Michael (2003): From Environmental Determination to Cultural-Historical
Mediation: Toward Biological Plausible Social Theories. Cybernetics & Human Knowing,
10 (2), 8-28.
Roth, Wolff-Michael (in press): Identity as Dialectic: Re/Making Self in Urban Schooling. In
Joe Kincheloe/Phil Anderson/Karel Rose/Derrick Griffith/Kecia Hayes (eds.), Urban
Education: An Encyclopedia. Westport, CT: Greenwood.
Roth, Wolff-Michael/McGinn, Michelle K. (1998): >unDELETE Science Education:/lives/
work/voices. Journal of Research in Science Teaching, 35, 399-421.
Rousseau, Ronald/Zuccala, Alesia (2004): A Classification of Author Co-Citations: Definitions and Search Strategies. Journal of the American Society for Information Science and
Technology, 55, 513-529.
Sewell, William H. (1992): A Theory of Structure: Duality, Agency and Transformation.
American Journal of Sociology, 98, 1-29.
Suchman, Lucy A. (1987): Plans and Situated actions: The Problem of Human-Machine
Communication. Cambridge: Cambridge University Press.
Turnbull, David (1993): The ad hoc collective work of building gothic cathedrals with templates, string, and geometry. Science, Technology, & Human Values, 18, 315-340.
Vygotsky, Lev S. (1978): Mind in Society: The Development of Higher Psychological
Processes. Cambridge: Harvard University Press.
Prof. Dr. Wolff-Michael Roth
Lansdowne Professor Applied Cognitive Science
MacLaurin Building A548, University of Victoria
Victoria, BC V8W 3N4, Canada
[email protected]